A long time ago, in a galaxy far, far away…in the late 20th century…a young chap started out in Computer Science at the University of Colorado in Denver.
Things were going pretty well. Second semester hit, and the adventure really began. There was a course in object oriented development in C++ in this lad’s early CS career. He hadn’t seen objects before, and only had a notional understanding of what they were—if even that.
A project that took pretty much all of the semester was to create a set of classes to manipulate and display a linked list. This was his first foray into abstract data types. It was quite the effort across that semester, but it got done—and, not only was a program which handled CRUD for a linked list the end result, but an indexed one—to allow you to access values non-sequentially as if it were an array.
It was a cool project in the end, and this lad looks back with fondness at those early days, when all of it was so new.
Flash forward nearly three decades to the advent of LLM, Large Language Models.
It’s become a bit of a hobby of mine to throw things at ChatGPT, one of the more popular LLM generators these days, to see what it can do.
I gave it the basic requirements of that very same semester project freshman year. It generated a whole class, complete with a main driver, in mere seconds. You can’t help but feel a mix of fascination and a bit of melancholy as you watch all of that effort (which I could now do in less than an hour, myself) do it all in just a mere moment.
Also makes me wonder how creative professors have to get these days to come up with a project that can’t be easily generated 😂